In order to prepare a node to install Hadoop , you can either install Linux on Virtual Box or VMware Player /VMware Fusion .
Let follow the steps to install the Linux OS on Virtual Box.
You can download the current version of Virtual box for Linux ,Window or Mac as per your requirement HERE .
The current version of Linux ubuntu can be downloaded from HERE .(I personally prefer the server version.)
Let’s start with a new virtual machine creation on the Oracle VM virtual box that you should have already installed following the instruction from Virtual box download link.
Provide the name to your machine and select the create a new virtual hard disk. As we are installing a Linux OS ,go ahead and select the Linux type and the version that you are going to install. Click Create.
Here you are creating a fixed size virtual Hard disk . 8 GB the minimum suggested RAM for any node on Hadoop cluster. You can increase the size for better performance. Click Create and the system will create the fixed storage medium.
The previous step creates your node but you still have an empty node without any Operating system.
Select the Node that was created and go into the Settings.
Select Storage.you should find that the controller : IDE is empty .
Select the drop down and from the folder choose the iso you would have downloaded from the Linux website earlier.(In my case , I have installed the ubuntu server 64 bit version.).Click Ok.
Start the node and follow the installation steps to install the OS in case of real hard drive.
Provide the Node with a hostname and also an alternate user name than the root.
As I am going to install MapR on this node , a mandatory step is creating mapr user account.
Test logging in the node.
Also make sure that you test that the IDE is empty again so that it does not load from the CD/DVD device or iso again and again,